Housing cycles and financial stability – the role of the policymaker
Speech by Peter Praet, Member of the Executive Board of the ECB,
at the EMF Annual Conference 2011,
Brussels, 24 November 2011
Housing markets have certain characteristics that intrinsically link them to financial stability. House price collapses can have systemic consequences. We know this from the current crisis and we know this from past crises. This fact raises a number of important questions for policymakers.
Do asset price bubbles exist?
If they do, how do we identify them?
Should we react in the same manner to all asset price bubbles? Should we treat housing price bubbles differently?
We know the cost of not reacting to house price bubbles, but what is the cost of over reaction
If we react, how do we react and who reacts?
These are all important questions and one could easily devote an entire speech to each of them. Instead, I’ll try to touch lightly on all of them.
Do asset price bubbles exist?
This may seem like a strange question to ask in light of the recent experiences in the US, Ireland and Spain. But, there is no consensus among economists that asset price bubbles actually exist. If one takes the narrow definition of a bubble used by researchers then – in order to identify a bubble – policymakers need to prove that, given the information available at the time, investors behaved irrationally. This is an impossible task, even ex-post. Some well reputed economists have argued that all famous historical asset price bubbles - from the Dutch Tulip Mania of 1634 to 1637 to the new economy boom of the 1990s - can be explained by fundamentally justified expectations about future returns on the underlying assets. Therefore, even though the booms were large and ultimately costly they were not considered excessive or irrational and, therefore, were not bubbles. So there was no need to react? I’m not trying to argue here that asset price bubbles don’t exist, I believe they do. But while the costs of dangerous house price bubbles are clear, their identification is not. Even if one feels certain that a bubble exists, it’s impossible to prove and there will many who will say you are wrong (this time it is different).
The systemic importance of house price cycles compared with other asset price cycles
The potential risks of asset price bubbles tend to vary across asset classes. Research by the IMF has shown that housing busts are, on average, twice as costly in terms of output losses as equity price busts. This reflects the higher exposure of banks to mortgages than shares. A key conclusion of literature on historical financial crises is that they tend to have worse outcomes when banks are distressed. For example, in October 1987 stock markets around the world fell sharply, but this did not represent a threat to the banking system and so, its impact was contained. In contrast, many researchers trace the origins of the current financial crisis back to the bursting of the US house price bubble. The link between house prices, banking crises and recessions is not unique to the current situation: a number of studies examining crises in both advanced and emerging economies over time and across countries have shown that they tend to coincide with the bursting of house price bubbles. The securitisation of mortgage loans has not only tended to weaken the origination process but has tended to expose countries’ financial systems to real estate excesses in other countries. By transferring the credit risk the originator can exacerbate the boom without fear of the consequences of a collapse.
Identifying house price bubbles: problems and potential indicators
So, how do policymakers identify these costly house price collapses? Most house price busts are preceded by lengthy booms, so perhaps policymakers should try to identify potentially costly booms. But, even identifying a house price boom (costly or not) in real time can be challenging due to data issues regarding house price indices. The quality of these indices can be quite poor, one country can have a variety of indices from which to choose (each providing a different measure of growth), the publication lag can be quite significant in some countries and, within the euro area, there is a lack of harmonisation across countries.
Even if one identifies a boom, how do you know that it’s a costly one? Not all house price booms end in bust and not all busts are very costly in terms of output nor have implications for financial stability. A 2003 IMF study of real house price cycles for 14 countries over 30 years found that only 40 per cent of house price booms ended in bust. An examination by the OECD of 17 countries over a similar period showed that two-thirds of real house price booms ended in bust. Closer to home, a 2003 ECB study of real house price cycles in EU countries over twenty years found that busts followed 55 per cent of booms. If we examine nominal house price cycles the percentage of booms ending in bust is even lower.
But, to avoid overreaction, policymakers need to be able to disentangle the pernicious housing price booms from those that are costless or low-cost. This is a difficult task, even ex-post, and one that some researchers have described as “impossible” ex-ante. However, research on house price booms and busts has identified some useful and timely indicators. Analysis has shown that house price booms fuelled by excess credit growth tend to be most costly. Following from this conclusion, research - within ECB, the BIS and the IMF – has shown that simple deviations of money and credit aggregates from a trend that exceed a given threshold are useful predicators of potentially costly boom/bust cycles in asset prices. A key feature of these indictors from a policy perspective is that they signal a warning well before the standard conjunctural analysis does.
Proactive and reactive policy approaches
Once there is a clear indication or signal that a house price is likely to be a reflection of a potentially costly boom episode, how should policy makers react? Policy measures aimed at containing asset price bubbles can be proactive and reactive. In the presence of financially sound borrowers and/or a resilient banking sector, the impact of house price declines on the real economy should be limited. This enhances the case for a proactive approach focusing on both the quality of the borrowers and the resilience of the banks.
Regarding borrower quality, it is imperative that banks ensure that mortgage holders are credit worthy and capable of withstanding income shocks. The propensity for banks to take risks can increase in a competitive market where they are trying to maintain/increase market share. It is worth noting at this point that the significant and widespread mistakes made by (sub-prime) mortgage lenders in the US when assessing the credit worthiness of borrowers were not paralleled in the euro area and that write-offs on housing loans extended by euro area banks have remained low and broadly stable this year. However, certain euro area mortgage lenders did make some serious errors in judgement and credit risk exposures arising from mortgage lending do vary significantly across euro area countries. Exposures are highest among banks operating in countries with high household indebtedness, subdued household income prospects and/or where there is potential for a decline in residential/commercial property prices and also among banks that issued loans in foreign currencies that have now appreciated. The tools available to limit the exposure of banks to borrower credit risk are well known by this audience but I’ll list a few anyway: loan-to-value thresholds, debt-to-income thresholds, caps on the monthly repayment-to-income ratio.
As regards the banks resilience, the current financial crisis revealed that fragmented micro-prudential supervision was insufficient to ensure a stable banking sector. Since then some substantial changes have been made and are currently under way in financial regulation. These include a strengthening of micro-prudential oversight and, perhaps most importantly, the introduction of a complementary macro-prudential perspective.
The role of the ESRB
The practice of macro-prudential oversight – the bird’s eye perspective on the financial system – was not yet sufficiently established in the period before the crisis erupted. In Europe a key response to the crisis was the creation last year of the European Systemic Risk Board (ESRB) – an independent EU body responsible for the macro-prudential oversight of the Union’s financial system.
An important challenge for the ESRB in its task of monitoring systemic risks and vulnerabilities is bringing this new macro-prudential perspective to the traditional micro-prudential one. This interaction is key, as the recent financial crisis painfully illustrated that financial institutions may be sound on a stand-alone basis while the financial system as a whole can still be exposed to serious risks and vulnerabilities. In the ESRB this interaction is facilitated by the body’s composition which brings together representatives from both central banks and financial supervisory authorities from all 27 EU Member States, from the European Commission and the three European Supervisory Authorities.
To prevent and mitigate systemic risks to the EU financial system, the ESRB identifies and assesses risks and vulnerabilities and may as a next step issue risk warnings and/or recommendations if deemed appropriate. Warnings and recommendations can be either public or private. They can be addressed to supervisory bodies, both national and European ones, as well as individual member states and the EU as a whole. The ESRB gives the addressee of a recommendation a timeline for its implementation, and this process is then followed-up according to a “comply or explain” mechanism. Although the ESRB has no binding power, addressees that do not follow recommendations have to provide reasons for their inaction. This mechanism, together with the “moral suasion” of a public recommendation, provides a strong incentive for the addressees to comply.
Last month, the ESRB published its first public recommendations on lending in foreign currencies. Lending in foreign currencies has become common in some EU countries. Often it takes place through mortgage loans at significantly lower interest rates than similar loans in national currency and to households that are not protected against the exchange rate risk. It entails significant risks for the financial sector, including the potential for cross-border contagion. It is a practise that may also amplify booms and busts in housing markets. The ESRB’s recommendations aim to tackle the roots of the problems and thereby, improve the overall resilience of the financial sector.
A lot of the ESRB’s work to date has, by necessity, focused on the challenges of the current crisis situation, but much effort has also gone into more medium term or structural issues and the development and future implementation of macro-prudential tools. Having the proper regulatory framework, the right tools and the mandate to use them is essential for achieving more balanced future developments, for example in credit and asset markets. The ESRB, therefore, reviews draft EU financial legislation, and provides suggestions to legislators on how to strengthen its macro-prudential dimension whilst ensuring that policy-makers have the flexibility to take necessary action.
The macro-prudential toolkit
The macro-prudential policy toolkit is currently in the process of being built. The potential instruments are various and many are already assigned to other policy objectives, most notably short-term interest rates, capital ratios and tools for ensuring the soundness of individual banks. Few of these instruments have been used for macro-prudential purposes; therefore, practical experience is limited. The choice of instrument will largely depend on the policy goal. If the aim is to constrain excessive credit growth, then tools that impact the cost and quantity of credit are the most likely choice. Given the key role of interest rates in achieving the ECB’s price stability objective, capital or leverage requirements, and/or liquidity requirements may be the most appropriate tools. If the objective is to curb borrower indebtedness, then the aforementioned tools are also useful, as are other measures including increasing collateral requirements (e.g. LTV ratios), changing the tax treatment of mortgage interest and capital controls (but risk of long-term distortions). If the goal is to reduce banks’ exposure to asset price cycles then counter cyclical capital buffers are useful as well as tools of a more micro-prudential nature.
Basel capital and liquidity standards
In addition to the establishment the ESRB, several steps have been taken both at the international and at the European level to improve the regulatory framework, by introducing financial regulations that directly or indirectly address systemic risks. These include the new Basel capital and liquidity requirements published in December 2010.
While the higher minimum capital requirements are aimed at strengthening micro-prudential supervision, they will be complemented by the introduction of counter cyclical capital buffers from 2016 onwards. The objective of countercyclical buffers is to ensure banks build up capital buffers in good times, so that capital is available to absorb increased losses in downturns. This tool is largely macro-prudential and should help mitigate excessive credit growth and avoid the build up of excessive risks – not just in mortgage markets - but in the financial system as a whole. Although the trigger for activating the counter-cyclical capital buffer is defined as the excessive deviation of private sector credit growth from its long term trend, authorities may use complementary sources of information in buffer settings, including more granular data on house prices and mortgage loans.
The new liquidity rules of Basel III should also mitigate excessive credit growth. In particular, the so called net stable funding ratio, which aims to reduce the maturity mismatch by requiring banks to finance their long term loans with long term funds. The Basel III framework, represents a minimum standard, and therefore, authorities can introduce more stringent prudential rules to address specific risks in their domestic banking systems, for example higher risk weights for certain exposure classes. For example, the proposed Capital Requirements Regulation (Art. 119), which will be directly applicable across all European countries, offers a leeway for national authorities to increase risk weights or set stricter criteria (e.g. for loan-to-value ratios) for exposures secured on residential or commercial real estate on the grounds of financial stability.
The role of central banks: Leaning against the wind?
The financial crisis has intensified the debate as to whether monetary policy should be actively used to contain asset price booms and bubbles. This is the idea that central banks should “lean against the wind” of a sustained and swift upward movement in asset prices. In general, the view on “leaning against the wind” in the central bank community seems to have shifted more in its favour since the outbreak of the crisis. In the past, there were some very vocal critics of “leaning against the wind” whose arguments included the following:
Monetary policy is too blunt a tool to contain bubbles. Increasing the main refinancing rate has implications not just for mortgage rates but also rates on loans given to other sectors of the economy and other potentially non-booming asset prices.
“Leaning against the wind” is a very complex tool to use in a currency union such as the euro area where one can experience diverging trends in house and other asset prices across countries
Finally, as I have mentioned, asset price booms are not always costly and it’s difficult to identify those that are in real time.
Since the outbreak of the financial crises, more and more empirical evidence and theoretical arguments have been put forward either directly or indirectly supporting the “leaning against the wind” proposition. The arguments supporting this proposition include the following:
The current and past crises highlight the potentially significantly high costs of asset price booms and busts
Simple deviation indicators help with the early detection of potentially costly house price booms/busts. (Deviation from long term trends in credit-to-GDP ratio is a trigger for the activation of the counter-cyclical capital buffer. Data on house prices and mortgage loans may also be used by authorities as complementary source of information in buffer settings.)
There is increasing evidence that banks’ attitude to risk are correlated with the monetary policy stance (risk taking channel)
Small increase in rates might break herding behaviour of private investors.
The debate is ongoing. Regarding the ECB’s stance, some elements of “leaning against the wind” are implicitly incorporated in the monetary policy strategy through the second pillar of monetary analysis. Having said this, it should be clear that “leaning against the wind” is only a second best instrument, to be considered once macro-prudential tools should have proven to be ineffective and only to be used in very special circumstances, e.g. in case a rise in asset prices is geographically widespread and thus threatening price stability in the monetary union in the long-run.
Asset price bubbles do exist and have potentially large costs for the real economy, particularly in the case of house price bubbles. Although their costs are clear their identification is not. The ex-ante detection of costly house price booms is extremely difficult and cannot be done with complete certainty. In addition, identification problems are amplified by data issues regarding house price indices. However, a number of useful predicators of potentially costly boom/bust cycles have been identified by researchers, these include simple deviations of money and credit aggregates from a trend that exceed a given threshold. Policy makers can use both proactive and reactive measures to combat costly house price booms. Limiting credit risk exposures from borrowers and/or ensuring the resilience of banks would help limit the potential costs of house price collapses. The establishment of the ESRB – an independent body responsible for the macro-prudential oversight of the EU financial system – along with the new capital and liquidity measures for banks marks a big step in the right direction. However, the macroprudential policy toolkit is currently in the build-up phase. The current set of instruments is varied and many are already assigned to other policy objectives. The instruments policy makers use will largely depend on the objective they want to achieve. The use of macro-prudential instruments requires a balancing act: on the one hand containing developments judged to contain financial stability risks and on the other hand ensuring that economic growth and financial development is not unnecessarily constrained by an overreaction. As regards the monitoring of asset price cycles and mitigating the risk of credit-fuelled booms, the newly formed ESRB, which brings together central bankers, financial supervisors (both national and European) and the European Commission, is best placed to provide warnings and recommendations to policymakers. We will never be able to create a “bubble free” world but, with better support from macro-prudential policy, we can significantly reduce the probability of costly booms.
Alessi, L. and C. Detken, “Quasi real time early warning indicators for costly asset price boom/bust cycles: A role for global liquidity”, European Journal of Political Economy, 27(3), 520-533, Sept. 2011.
Allen, F. and K. Rogoff (2011), Asset Prices, Financial Stability and Monetary Policy, paper presented at Swedish Riksbank Workshop on Housing Markets Monetary Policy and Financial Stability (subsequently updated), February.
Clark, A. and A. Large (2011), Macroprudential Policy: Addressing the Things We Don’t Know, Occasional Paper Series, No. 83, Group of Thirty, Washington.
Drudi, F. et al. (2009), Housing Finance in the euro area, Occasional Paper Series , No. 101, European Central Bank, March
ECB (2003), Structural Factors in the EU Housing Markets, European Central Bank.
Garber, P. (2000), Famous first bubbles: the fundamentals of early manias, Massachusetts Institute of Technology.
Girouard, N., M. Kennedy, P. van den Noord and C. Andre´, (2006), Recent house price developments: The role of fundamentals, OECD Working Paper, No. 475, Organisation for Economic Co-operation and Development.
IMF (2004), “When Bubbles Burst”, World Economic Outlook, 61-84, International Monetary Fund, April.
Pastor, L. and P. Veronesi (2004), Was there a NASDAQ bubble in the late 1990s?, Working Paper, University of Chicago.
Reinhart, C. and K. Rogoff, Banking Crises: An Equal Opportunity Menace, Working Paper Series , No. 14587, National Bureau of Economic Research, December.
I would like to thank Nicola Doyle, Petra Lennartsdotter and Balázs Zsámboki for their contribution to the preparation of this keynote speech.